“No man is an island,” wrote John Donne. And yet, the development of big data has enabled companies to better see the differences between customers. “A piece of the continent, a part of the main,” we may be, but the explosion of information in the era of the internet and smartphones means that – in product and marketing terms – we can be treated as islands. Or at least as archipelagos of similarities.

Evidence for the impact of big data is legendary. Granular insights about how we use and move through the world shape our built environment and transport networks. Medical data helps to keep us well by advancing new treatments or eking out healthcare system efficiencies. Big data provides intelligence that guides or even automates business decisions. And, at the end of a long day, big data is the source of predictions that curate music and video libraries to our individual tastes.

Big data and analytics have become indispensable for product leaders in many industries – driving product innovations, improving competitiveness, and enhancing the customer experience. We’ve seen this in many parts of the lending industry. Digital data about credit histories, transactions, employment and location, combines with financial trends to inspire new products and tailor interest rates to serve the needs and desires of highly specific customer segments.

Product innovation has radically altered the experience of borrowing for consumers. From auto loans to credit cards, choice has mushroomed and approval times are usually counted in minutes. Since all or most of the data is digital, underwriting processes move incredibly fast. If you’re in the market for a new credit card, it can all be completed in a matter of minutes. Some lenders even provide an electronic card that works from your smartphone, enabling you to use your new line of credit straight away.

In comparison, the experience of securing a mortgage – by far the biggest category of lending– is stuck in the slow lane. Most homebuyers find that, unlike the minutes or hours to complete other types of loans, initial underwriting approval for a mortgage typically takes several days once a full loan file has been submitted.

Borrowers may be frustrated, but this isn’t mortgage providers’ fault. Lenders are constrained by the need to directly review borrowers’ documents. Pay stubs, W-2s, tax returns, property appraisals, and more – all need to be physically checked and classified to ensure the file is complete.

That’s just the start of the process. Invisible to borrowers is the work being done to extract information from documents and turn it into the digital data that can be fed into loan origination systems (LOSs).

Human error in data extraction means that data and the underlying documents often need to be checked and rechecked at each stage of loan manufacturing process. This means that mortgage “turn times” – the time it takes from underwriting to closing – can run to several weeks.

Mortgage turn times are slow even when everything is in order. Times can be extended further when application volumes are high, staff numbers are low, or there are errors and complexities in borrowers’ profiles.

Mortgage product teams lack the granular detail available to other categories of lending, which hinders their ability to innovate, identify competitive advantages, and improve market share. Compare mortgages with the proliferation of products and slick customer experiences in credit cards and personal loans and the evidence is clear.

Curbs to innovation and the customer experience are opportunity costs – hard to quantify but no less real or significant. As mortgage companies struggle to develop more competitive and diversified products, or frustrate borrowers with slow service and time-to-close, they limit their business potential.

The results are lost sales, hampered growth, and possibly bad customer reviews. When these circumstances persist, there can be lasting damage to the mortgage company brand and a lack of business resilience in the face of changing market conditions.

Technology for reading borrowers’ documents has, until recently, struggled to match the capabilities and accuracy of trained people. However, task-focused artificial intelligence (AI) now offers an alternative. The latest systems are easily able to outperform people in both speed and accuracy on docs-to-data conversion and data verification tasks.

Operations teams are integrating these capabilities with LOSs, streamlining mortgage manufacturing to achieve faster throughput with reduced risk. Alongside the gains in data speed and accuracy, AI also expands the scope of the information can be extracted.

Not long after productivity and efficiency gains are seen in operations, product teams realize that they have an untapped resource of incredibly rich, highly accurate and fully digital borrower data. This can be combined with other forms of digital data to turbocharge innovation.

For product teams, the transition to AI holds the potential to overcome opportunity costs and gain the digital data advantage that will enable mortgage industry innovation to finally compare with that seen in other categories of lending.

5 Ways Lending Intelligence Makes Life Better for Product Teams

Product leaders in the mortgage industry are always seeking ways to improve their business model and gain market share. Over recent years, some of the industry’s biggest gains have been achieved through LOSs. Although these systems have helped streamline loan manufacturing, the same cannot be said about the borrower data that powers them.

The work of converting borrowers’ documents into data that LOSs can use remains highly manual and expensive, restricting lenders’ ability to react to fluctuating demand.

That no longer needs to be the case. TRUE Lending Intelligence solutions finally break the link between workforce staffing and tasks such as document classification, data extraction, and data verification. That unlocks huge potential for product teams.

It’s fair to say that there have been some exaggerated claims about AI by some vendors, but TRUE’s AI is uniquely built for the lending industry – it is entirely focused on enabling mortgage lending, servicing, and insurance businesses to achieve breakthrough performance. Here are five ways TRUE is helping mortgage industry product teams.

Improve customer knowledge

Every lender needs to determine the credit worthiness of an applicant. For smaller sums, ranging from credit cards to auto loans, there is a wealth of digital data that enables underwriting to be highly automated. Many of these loan decisions are made in minutes.

Mortgage providers use the same data, but it’s not sufficient to satisfy the larger sums involved in home purchases or to meet “qualified mortgage” regulations for lender protection and secondary market trading. Satisfying these requirements involves a document review that takes in income, assets, outgoings, and the property appraisal.

Documents reveal so much more about a borrower than universally available digital records, hence why they are essential for determining a borrower’s ability to repay the mortgage. However, classifying documents, then reading and extracting the relevant data, takes time and expertise. It’s a largely manual process, which makes it costly, so most mortgage firms limit their focus to what underwriters require. The rest of the information is often overlooked.

With TRUE, classification and extraction processes are fully automated. That enables mortgage companies to extract more data and build richer borrower profiles with no extra cost. This is achieved with levels of speed, accuracy and completeness that comfortably exceed anything achieved by human teams.

Rich and high-fidelity customer knowledge helps product teams to reduce risk and be more competitive with lending rates and fees. This knowledge can make a valuable difference to margin, protecting the business in leaner times and providing the foundation for growth when volumes are higher.

Turbocharge innovation

If you’re in the market for a credit card, you’ll find endless options: elite cards with exclusive benefits, cards that reflect your interests and values, specialist cards for those with poor credit histories.

Product innovation has driven credit card lending since Sears pioneered payment cards in 1911. Did you know that the earliest airline-branded cards date back to the 1930s? Or that introduction of “revolving credit” cards in the 1940s (allowing customers to make fresh purchases while they still owed money) paved the way for today’s general purpose and universally accepted cards.

While digital data has turbocharged innovation for some categories of lending, mortgage providers have been held back by the need to process large volumes of unstructured data. Digitizing information in borrowers’ documents swallows significant resources, limiting time and information available to product teams.

With docs-to-data and verification processes automated by TRUE, these restrictions disappear. Indeed, by combining universally available big data with proprietary borrower data extracted from documents, mortgage providers can develop a unique and exceptionally detailed source of customer knowledge.

Deep analysis reveals patterns, opportunities and risks that can be the basis of a refreshed product strategy. Product-to-market fit can be substantially improved, enabling better targeting of specific customer segments and greater success in winning market share.

Enhance customer experience

No one relishes the prospect of a mortgage application, but AI-powered document classification can make the task much easier for borrowers and dramatically speed up initial underwriting and turn times.

Lenders’ point of sale (POS) systems tell borrowers what documents they need to provide. The latest include easy-to-use apps, allowing documents to be photographed and uploaded in moments. However, behind the technology, these systems rely on people checking that applications are complete and documents have been correctly classified. Fixing problems – missing, mislabeled, or poorly imaged documents – can take several days.

TRUE can integrate with POS systems, instantly checking documents as they appear and advising borrowers of any problems. The time needed to confirm a complete application file is reduced from days to minutes. Lenders can immediately move to data extraction and begin the mortgage manufacturing process.

Accuracy and completeness in borrowers’ profiles means the entire manufacturing process can be sped up. This makes it possible for mortgage providers to compete not only on rates and fees, but also service speeds and digital experience. These factors are becoming increasingly important as Gen Y and Z consumers, bringing high expectations of digital financial services, take a larger share of the homebuying market.

Improve the business model

Most human effort is expended at the start of the origination process: documents classification and data extraction take trained agents many hours per mortgage application.

Lenders have made significant investments in LOSs, many of which include automation technology intended to reduce human input into docs-to-data processes. Despite this, manufacturing costs per-loan have remained stubbornly high. Why?

The problem is a lack of accuracy and completeness. Lenders find that many supposedly automated systems need significant human intervention to spot and correct errors. That undermines a key part of LOS value and limits cost savings. It’s typical for a 5 percent increase in docs-to-data conversion errors to multiply to a 30 percent increase in human intervention later in the lending process. The later errors are noticed, the more it costs to resolve them.

As the only purpose-built, AI-powered lending intelligence system on the market, TRUE delivers unprecedented data accuracy, starting at 95 percent out-of-the-box and exceeding 99 percent within months. This significantly reduces human intervention while radically improving speed and efficiency.

Such outcomes allow product leaders to rethink business models, using cost savings to reorganize resources, develope products for greater competitiveness, and focus on market share.

Increase business elasticity

Adapting to cycles of falling and rising demand is a recurring challenge. Many product leaders agree that a quieter period is the perfect time to make changes to business processes. As mortgage applications start to pickup, so does the demand on processing, which can be alleviated with the power of AI.

Supplementing loan origination and servicing processes with AI that fully automates docs-to data conversion and verification tasks is a game-changer. Data quality improves dramatically as costs decline. Loan manufacturing accelerates, risk falls, and there’s less need for hiring or outsourcing when demand returns.

In the short term, operating costs decrease. Over time, more benefits emerge: improved customer satisfaction, greater flexibility to adapt to changing market conditions, insights for product innovation, and the data foundation necessary to achieve the long-vaunted goal of full touchless automation.

Product teams in the mortgage industry are challenged to develop new products that are profitable and competitive. However, loan rates are far from the only factor that determines customer demand and satisfaction. According to research by McKinsey, as many as 58 percent of residential mortgage customers are dissatisfied with the lending process.

Product teams now have an opportunity to reassess their strategy and ensure they’re delivering an optimal selection of mortgage products and the best possible service.

Empowering lending professionals to help customers

In an age where competitors are always a quick Google search away, companies are paying more attention to the quality of their customer experience.

According to the McKinsey report, borrowers want mortgage providers to make the lending process simple and fast. That means getting all the right documents uploaded, with rapid advice back to borrowers if anything is missing, incorrect, or when additional documentation is needed. Meeting these expectations is essentially impossible when data extraction and verification processes are manual.

With TRUE’s AI powering docs-to-data and verification processes, TRUE customers are analyzing incoming borrower documents in a matter of minutes. TRUE helps underwriters check that all documents are in place, flagging any issues. Loan officers and automated messaging systems can step in to nudge borrowers to provide whatever is needed to progress their application.

Doing more with less

When volumes are high, mortgage providers tend to stick to the easiest type of loan since there’s limited elasticity in the business model; manual operations processes mean that throughput can’t scale easily or quickly. Having more business than you can handle sounds like a high-class problem, but product leaders know that it limits opportunities to win new borrowers and grow the business.

TRUE dramatically increases elasticity for lenders – automating critical processes with astounding accuracy at unprecedented speeds. With TRUE, product teams can be confident that their business can scale to demand while reducing costs even as volumes fluctuate.

Navigating leaner times

When the housing market slows down and interest rates rise, volume decreases. Product teams often respond by developing alternative offerings that keep the business moving and reduce idle time for loan officers and underwriters.

Low volumes can tempt mortgage providers to alter risk profiles and accept borrowers that they might not normally have favored. For example, bank statement or stated income mortgage products may be introduced, which could potentially lead to riskier customers.

In these market conditions, more detailed and accurate data can mean the difference between a good or bad lending decision.

Optimizing for product quality

One of the drags on the introduction of new products is the need to train loan officers and underwriters on changed processes. Although this work is handled by operations teams, product leaders need to be confident that new products can be brought to market with consistent quality for risk analysis and the customer experience.

Innovative products – perhaps introduced as the business adapts to changing market conditions – come with a heightened risk of errors. A client launching bank statement mortgages used TRUE to automatically capture all the data underwriters would need from each bank statement, even when statement designs varied. This avoided the need for underwriters to analyze bank statements manually and allowed more time for dealing with exceptions. By providing reliable and consistent docs-to-data conversion and verification, TRUE made it easier, faster and less risky to bring a new product to market and be confident in the quality of loan decisions and the customer experience. Knowing that loan data will be clean and complete also makes it easier for lenders to package up loan books for the capital markets – with reduced risk of having to repurchase loans due to errors.

Helping product teams reach their full potential with TRUE

Today’s mortgage market means many lenders urgently need to improve their products and customer experience but can no longer afford to simply add more people. These pressures are leading smart lending businesses to look afresh at automation technology that can take care of key data processes – reliably and at scale.

The outcomes are richer and more accurate borrower data, and more capacity in both product and operational teams. This leads to more resilient and innovative businesses that are equipped to deliver a better customer experience.

Reimagine your Mortgage Products and Customer Experiences with Lending Intelligence

Lenders are using AI to capture richer and more accurate borrower data, leading to new product and customer service innovations.

Mortgages Stuck in the Slow Lane

Frustrated Potential for Product Teams

Gain the Digital Advantage with Lending Intelligence